28 May 2026 08:52 PM
Hi community,
I'm playing around with the new user sessions and user events. I'm finding out that I've got multiple sessions with the same dt.rum.session.id.
My question: what makes a session unique and how does Dynatrace correlate the right user events to the right session? I can't dicover this using fetch user.events and correlating it with user.sessions etcetera.
Anyone got some tips? 🙂
02 Jun 2026 02:58 PM
Hi,
I tested this myself and found something interesting. A dt.rum.session.id visible in user.events is not necessarily present in user.sessions.
fetch user.events
| filter dt.rum.session.id == "VHDBVRRUSPKKUSFPCCMRNUSRHTSWESHQ-0"
| sort timestamp ascreturns events, while:
fetch user.sessions
| filter dt.rum.session.id == "VHDBVRRUSPKKUSFPCCMRNUSRHTSWESHQ-0"
| sort timestamp ascreturns no records.
This suggests that user.sessions is not simply a grouped view of user.events. There appears to be an additional aggregation/correlation layer that creates session records asynchronously and possibly applies extra conditions before a session becomes available.
So, does anyone from Dynatrace know:
03 Jun 2026 11:39 AM
Hi!
I'm Ramon, part of the team in charge of this 🙂
> what uniquely identifies a record in user.sessions?
A combination of `dt.rum.session.id` and `start_time`
> how exactly user.events are correlated into user.sessions?
For all events happening within a specific timeframe that contain the same `dt.rum.session.id`, they'll be aggregated into the session
> whether dt.rum.session.id is expected to be unique across user.sessions?
Sadly, no. It's unique in time for a specific timeframe (there can't be two sessions with the same id at the same time), but it's definitely not unique.
Feel free to drop more questions if you have any 😄
Featured Posts